Plant genotypic diversity does not beget root-fungal species diversity
- First Online:
- Cite this article as:
- Johnson, D., Anderson, I.C., Williams, A. et al. Plant Soil (2010) 336: 107. doi:10.1007/s11104-010-0452-9
- 162 Views
The number of genetically distinct individuals within a community is a key component of biodiversity and yet its impact at different trophic levels, especially upon the diversity of functionally important soil microorganisms is poorly understood. Here, we test the hypothesis that plant communities that are genetically impoverished will support fewer species of root-associated fungi. We used established grassland mesocosms comprising non-sterile natural soil supporting defined communities of 11 clonally-propagated plant species. Half of the mesocosms contained one genotype per species and half 16 genotypes per species. After 8 years growth, we sampled roots from the mesocosms and measured root-associated fungal richness and diversity using terminal restriction fragment length polymorphism (T-RFLP). Contrary to our hypothesis, we found that the roots of genetically impoverished communities contained more species of fungi and had greater diversity compared to genetically rich communities. Analysis of the plant species composition of the mesocosm communities indicated that genotypic diversity affects root-fungal diversity indirectly through its influence upon plant species diversity. Our findings highlight the need to include feedbacks with plant intraspecific diversity into existing models describing the maintenance of soil biodiversity.
KeywordsIntraspecific diversityGenotypic diversityArbuscular mycorrhizal fungiSpecies diversitySoil microorganismsEcosystem functioningBiodiversityTemperate grassland
The acceleration of species extinctions is such that understanding relationships between biodiversity and ecosystem function is a crucial aspect of ecology. Community composition is strongly influenced by the genetic diversity of its constituent individuals, for example, via competitive interactions, genotype by environment interactions and by genotypes affecting key ecological processes. We must therefore consider how variation in genotypes, and not just species in a community drives ecosystem functioning (Hughes et al. 2008).
The role of genotypic (i.e. intraspecific) diversity has recently been considered in experimental grassland mesocosms, where it was found that genetically impoverished plant communities became less diverse and more divergent in species composition than in communities comprising more genotypes of each species (Booth and Grime 2003). This finding is important because interactions between plants and other organisms can determine their impact on ecosystem functioning (Wardle et al. 2004). In temperate grassland, >80% of plant species always form associations with mutualistic arbuscular mycorrhizal (AM) fungi (Smith and Read 2008), the activity and diversity of which have crucial roles for host plant nutrition and productivity. The community structure of AM fungi in roots of co-occurring grassland plants has been shown to have non-random distributions (Vandenkoornhuyse et al. 2003), and there is evidence of specificity at the plant population level (Ronsheim and Anderson 2001). These observations support findings from a range of species interactions that argue ‘diversity begets diversity’. From this, it follows that plant intraspecific diversity could impact on AM fungal communities either directly or indirectly, by acting via changes in plant species diversity. Direct impacts on mycorrhizal fungi are most likely when there is large variation in traits that have a role in regulating fungal colonisation or functioning. In lightly grazed limestone grassland, variation in a range of important traits amongst plant genotypes has been demonstrated within a 10 × 10 m area (Whitlock et al. 2010).
Here, we test the hypothesis that genetic impoverishment of plant communities leads to a less diverse community of fungi colonising plant roots. We do this by utilising a long-term experimental manipulation of the intraspecific diversity of plant communities derived from a natural population and grown under constant environmental conditions.
Materials and methods
The experiment was established in 1998 (Booth and Grime 2003) and comprises thirty-six model plant communities (0.6 m × 0.6 m in size) assembled from experimental populations of 11 plant species, each population drawing upon a pool of 16 putative genotypes. The species used in the mesocosms were: Festuca ovina, Briza media, Koeleria macrantha and Helictotrichon pratense (grasses), Carex caryophyllea, C. flacca and C. panicea (sedges), Succisa pratensis, Leontodon hispidus. Viola riviniana and Campanula rotundifolia (forbs). The experimental genotypes were originally derived from established plants sampled at random from a 10 m × 10 m area of limestone pasture in Derbyshire, UK, and the genetic uniqueness of the genotypes verified (Whitlock et al. 2007). Each experimental community was initially identical in species composition by planting 16 individuals of each species in all communities. The communities were managed by cutting and by removing all flowers by hand to prevent plants from dispersing seeds back into their communities (Booth and Grime 2003). The relative abundance of the five most abundant species (which have the greatest influence on species diversity of the communities; Whitlock et al. 2007) was recorded using point quadrats. The abundance of each surviving genotype was measured using point-quadrat sampling of leaf tissue followed by genotype identification using molecular markers (Whitlock et al. 2007). These data were used to calculate Shannon indices of species diversity.
In the present study, we used a subset of mesocosms: eight mesocosms contained one genotype per species and five contained 16 genotypes per species. This was half the number of mesocosms available for each treatment, and these were selected randomly (but the identities of genotypes in the 1 genotype per species mesocosms were noted). In July 2006, two soil cores (2 cm diameter) were removed from each of the mesocosms. Within a mesocosm, each of the two cores were located in separate halves of the lysimeter box (i.e. they were spatially distinct). Plant roots were washed in sterile water and their DNA was extracted. The species composition of root-associated fungal communities was analysed by terminal restriction fragment length polymorphism (T-RFLP) using a nested PCR approach with primers that have high specificity for AM fungi (Gollotte et al. 2004). The primers LR1 and FLR2 (Trouvelot et al. 1999; van Tuinen et al. 1998) were used to amplify the 5’ end of the LSU rDNA and in the 2nd round PCR, the fluorescently labelled primer pair FLR3 and FLR4 were used (Gollotte et al. 2004). Purified PCR products were digested with restriction enzymes TaqI and MboI (Mummey and Rillig 2007). Digestions were analysed on an ABI PRISM™ 3130xl genetic analyser (Applied Biosystems, Warrington, UK) to detect and measure the relative abundance of individual restriction fragments. The numbers of individual peaks from the T-RFLP analysis were used as a proxy for species richness following established procedures (Johnson et al. 2004). Cloning and sequencing was undertaken to confirm the identity of 8 of the most abundant T-RFLP types; cloned sequences were subjected to a BLAST search (excluding sequences derived from environmental samples) in GenBank.
Root-associated fungal communities were characterised by both the number of peaks (richness) and their fluorescence intensity (corresponding to relative abundance) from the T-RFLP analysis based on established procedures (Johnson et al. 2004). Root fungal diversity was estimated using the Shannon index. Richness and diversity was computed separately for the two restriction enzymes and the average used to generate the final measures. Fungal richness data were analysed using general linear models with ‘soil core’ nested within the mesocosm treatment. All residual variances were checked for normality.
Here we show that plant genotypic richness in communities can lead to significant changes in the species composition of root associated fungi; it is likely that the majority of these species were AM. Contrary to our hypothesis, the diversity of root-associated fungi decreased in communities with the greatest number of genotypes per species. We predicted the opposite because we expected genetically rich plant communities to be analogous to a community with greater species richness and provide more niches for AM fungi to exploit.
It is difficult to ascertain unequivocally the mechanism through which plant genotypic diversity could regulate root fungal species diversity. There can be considerable variation in key physical and physiological traits within a plant species (Whitlock et al. 2010), sometimes of similar magnitude, or exceeding variation between species. While rarely considered at the intraspecific level, such variations between species are thought to be a key mechanism by which the activity and diversity of soil processes and organisms are regulated (De Deyn et al. 2008). In the context of the plants used in the present study, there is certainly a wide variation among genotypes in key traits including biomass, guerrilla—phalanx clonal growth strategy, relative biomass allocation between above-, below-ground and reproductive structures and many individual aspects of morphology (Bilton 2008; Whitlock et al. 2010). Further studies are needed to determine if our results can be explained by either the direct or indirect action of intraspecific variation in these traits upon AM fungal diversity and function.
Our observation that the diversity of the most abundant plant species in the genotypically rich treatment was less than in the 1 genotype per species treatment suggests that indirect effects may play a role in shaping AM fungal communities. In other words, plant genotypic diversity influenced species diversity, which in turn influenced fungal community structure. While the plant communities had identical species composition upon establishment in 1998, in the intervening period their species compositions have changed at different rates and with different trajectories, depending on their initial genotypic diversity (Booth and Grime 2003). Although we cannot unequivocally test this indirect mechanism here, a similar mechanism has been established in these mesocosms above ground, where the direct effect of genotypic composition has been shown to shape the species composition of the plant communities (Whitlock et al. 2007, 2010).
The observed responses could also be explained by particular plant genotypes providing more favourable hosts for a diverse array of fungi, but in mixture these hosts become subordinate. In our experiment there was little overlap in the identity of the genotypes that became dominant in the genetically rich mixtures with those in the genetically impoverished mesocosms. The only species where the dominant genotypes in the mesocosms containing 16 genotypes per species were represented in several (5) of the 1 genotype mesocosms was Carex caryophyllea (specifically genotype Cc09), but this species is non-mycorrhizal. This observation may indicate that genotypic diversity could act upon root fungal diversity through a change in dominance hierarchy of plant species, a situation that is predicted to occur between species of plants and AM fungi (Urcelay and Diaz 2003). The presence and identity of AM fungi has also been shown to affect the competitive balance between and within species in grassland (Moora and Zobel 1996). One possibility is that a highly clonal and physically large genotype of C. caryophyllea (genotype Cc09), which appeared in 4 of the 1-genotype communities we studied, might influence fungal richness and diversity through its competitive interactions with other (mycorrhizal) species present in the communities. Competitive genotypes of C. caryophyllea may interrupt the process whereby individual fungal species attain dominance, by diverting the resources of host plant species away from symbiosis and towards competition, e.g. for light, growth or other important traits.
We know from experimental manipulation that AM fungal species richness can have considerable effects on productivity and nutrition of plants in species-rich grassland (van der Heijden et al. 1998). Moreover, whilst there is theoretical and empirical support underpinning relationships between dominant and subordinate species of plants with AM fungi (Urcelay and Diaz 2003) and the role of these relationships in maintaining species diversity (Bever 2003), our data indicate that feedback mechanisms between plants and root associated fungi need to include variation at both intra- and interspecific levels.
This work was supported by the Royal Society and the Natural Environment Research Council. We thank Dr A.F.S. Taylor, P. Parkin and H. Weitz for their assistance. ICA received support from the Scottish Government.